Every manufacturer will tell you their technology is the right one. It depends entirely on what you are tracking, where, and what decision the location has to drive. Fluidity works across eight location technologies and holds relationships with more than twenty vendors — which is precisely why our recommendation is worth something.
No single technology suits every use case. That is the entire reason an independent recommendation is worth something — and the reason the answer changes from site to site.
In that order, every time. Accuracy is an output of these answers, not an input — and a specification written before these questions are asked is the most common reason a deployment disappoints.
People, tooling, stock, vehicles, work-in-progress, consumables.
Tag cost tolerance, form factor, battery life and the replacement workflow. At scale this is usually the real constraint — not the positioning spec.
One room, a whole site, a yard, a mine, indoor to outdoor, multiple locations.
Infrastructure density, whether GNSS is in play at all, and whether you need one technology or several stitched into a single picture.
“Which room is it in” versus “is it within 30 cm of that robot”.
Required precision — and this is where most budgets are wasted. Paying for centimetres to answer a room-level question is the most expensive mistake in this market.
Steel, racking, liquids, moving metal, wash-down, temperature, ATEX.
What will actually survive contact with your site. This is what kills projects specified from a brochure, and it is the reason we walk the building before we quote.
Which systems consume it, how fast, under what governance.
The integration surface, the middleware layer, and whether the data stays portable if you change hardware later.
Every location technology trades precision against coverage against cost. This is the trade honestly drawn — including the corner every vendor’s brochure claims to occupy, and nothing does. The full detail is in the table below.
Including where each one struggles. A vendor cannot publish this column — which is exactly why it is here.
| Technology | What it is genuinely for | Where it struggles |
|---|---|---|
| UWB | Tight precision where a decision depends on exact position — tooling, robot interaction zones, safety envelopes, collision avoidance. | Cost per square metre and infrastructure density. Comprehensively overkill for “which room is it in” — and it is an indoor technology; outdoors belongs to GNSS or radar. |
| BLE AoA | Sub-metre coverage across large continuous areas at a viable cost — the workhorse for wide-area people and asset tracking. | Locator geometry is unforgiving. A design that looks fine on paper can resolve angles badly in the real ceiling. |
| BLE | Presence, proximity and zonal awareness at the lowest cost per tag, on the largest device ecosystem in existence. | RSSI positioning is coarse — metres, not centimetres. Not for anything safety-critical. |
| RFID | Item-level identification at a chokepoint — gates, doorways, portals, production steps. Passive tags cost pennies, which changes what is economic to track at all. | Presence at a reader, not position between them. No continuous location, and read reliability is very sensitive to metal and liquid. |
| Wirepas | Massive scale and low power where running infrastructure everywhere is impractical. The network builds and heals itself. | Zonal rather than precise, and a genuinely different design discipline from anchor-based systems. |
| LoRa | Very long range at very low power across sites, yards and rural areas where infrastructure is sparse or expensive to run. | Metres at best, often more. Low update rates. Not a real-time precision tool. |
| GNSS / GPS incl. RTK | Everything outdoors. The only sensible answer for vehicles, yards, plant and wide-area logistics — with RTK where centimetres matter in the open. | Nothing indoors. Power hungry. Needs sky view, which yards and covered areas do not always provide. |
| Radar mm-wave | High precision across large environments — indoors and outdoors — with minimal infrastructure. A fundamentally different economic shape when an anchor grid is impractical: one radar covers what would otherwise take a grid of anchors, and the tags run for years on a single battery. | Newer and less proven at scale. Line of sight. A much smaller vendor ecosystem to fall back on. |
Most real deployments use two or more of these together. Stitching them into one coherent picture is the work.
Fluidity participates in the two bodies that determine whether location systems from different manufacturers can work together. That matters to you for one reason: it is what stops a location deployment becoming a permanent commitment to whoever you bought from first.
Governed by PROFIBUS & PROFINET International. omlox defines a common interface and coordinate model so position data from any system — UWB, BLE, Wi-Fi, RFID, GPS, 5G — arrives in one standard format. Every member holds an equal vote on how the standard develops.
Why you should care: mix and match hardware, and change vendors later without rewriting the integration.
FiRa writes the specifications that make ultra-wideband chips from different manufacturers interoperate, and runs the certification programme that proves it. Its members include Apple, Google, Samsung, Qualcomm, Bosch and NXP.
Why you should care: FiRa Certified is your evidence that UWB infrastructure will work with other vendors' equipment — and with the phone in your pocket — rather than stranding you in a proprietary dialect.
No single manufacturer covers every use case. We hold working relationships across the market — positioning platforms, device makers and data layers — which is what allows the recommendation to follow your requirement instead of our margin.
The locators, anchors and readers that go up across the space, and the positioning engines behind them. This is where the core architectural decision gets made — and the one that is most expensive to reverse.
The unglamorous decision that most often decides whether a deployment survives its second year.
Platforms that turn positions into decisions, and the specialists who solve a specific shape of problem.
A location system that lives in its own dashboard is a curiosity. Fluidity manages the integrations into the platforms your people already open every morning — operational systems that need to act on an event, and the BI and analytics tools where the accumulated history becomes something you can actually interrogate. These are the ones we work with most, and the list is not a limit.
Positioning hardware tells you a tag is at a coordinate. It has no idea that the coordinate is a quarantine bay, that the tag is a batch of finished goods, or that eleven hours is a problem. Phantom is the layer that knows.
Phantom is built AI-native — not a location system with a chatbot bolted to the side, but one where the meaning of a zone, an entity and an event is part of the data model itself. It is founded on Fluidity's patent-pending technology and on twenty years of finding out what location data actually has to do once it leaves the hardware.
Tell Phantom what to watch for in plain English. It shows you exactly what would have happened last week. Then you arm it — and when it fires, the paperwork writes itself.
Underneath: live floor-plan operations, dashboards, scheduled documents — and hardware-agnostic ingestion across whatever positioning technology is on site, with signed, filtered delivery into the systems you already run.
The pending patent covers the semantic logic at Phantom’s core: a structured layer of meaning — what a site is, what each zone is for, who or what each entity is, what counts as normal here — bound to the live position stream. It is what lets generative AI answer from the facts of your operation instead of inventing around them.
The same protected method keeps the AI honest in both directions: the vocabulary it writes rules and reports in is exactly the vocabulary the deterministic engine executes — and nothing it writes goes live until it has been previewed, proven against your site’s history, and armed by a human.
Good — that is the right place to start. The answer comes from your use case and your site, not from a price list.